Re: Austronesian lexical categories & voice (was: Trigger language question etc.)

Ray Brown wrote:
> On Sunday, May 15, 2005, at 06:00 , Roger Mills wrote:
> > In Indo., no matter how "noun-y" a word may be, most can at minimum take
> > the
> > ber- prefix, meaning "to have..." or "to do ...habitually or as a
> > profession" or "to have the quality of..." et.al.  It seems that in
> > Tagalog,
> > almost any "noun" can take the mag- prefix, which interestingly is
> > cognate
> > with the Indo. one, though I don't think the meanings are quite the
> > same.
>
> ber - and mag- are cognate? The /b/ ~ /m/ alternation is OK - but the
> /r/ ~ /g/ one is a bit odd. What is their derivation?
>
Actually it's the Ml/Indo. /b/ that's odd, but it's standard literary from
way back. Some peninsular Ml. dialects do have mer- (the "e" is schwa,
remember), Old Ml. (ca. 10th C) and other relatives have mar-.

The R/G correspondence is very well established (sometimes called the RGH
Law**) and was formalized almost immediately once real comparative work
began in the 19th C:  Ml. r, Tag. g, some langs. h, Jav. (and Polynesian) 0,
and it's currently symbolized as *R-- Dempwolff used "gamma", implying that
it was probably a velar fricative of some sort, and a resonant, not a stop
(since it never occurs with prenasalization). Other reflexes include y (and
> z), l (< r).
So the prefix is now reconstructed as *maR- (some have suggested *ma-R-, in
a system *ma- 'stative', *ma+N- 'active', *ma+R-(uncertain mng., perhaps
'habitual/frequentative'??) all verbal, along with *pa- 'causative', *pa+N-
'agent noun', *pa+R- (various functions, both nominal and verbal).

Another correspondence noted early on was the so-called RLD** Law-- Jav. r,
Tag. l, Ml. d --though it's much more complicated than that, and some even
believe that, due to rarity and distributional oddity (never in initial
position), the corr. probably ought to be aligned somewhere else. Dempwolff
called it "g-acute", modern *j; it was part of the stop system and
supposedly a palatal. Toba Batak and some Formosan and N.Phil. languages
reflect it as /g/ (while they reflect *R as something else), and it merged
with the other palatals in Oceanic.
-------------------
**Formerly called Van der Tuuk's 1st and 2nd Sound Laws (I forget which was
#1, which #2

Reply